Pricing¶
Autopoetic plans are weekly, and every plan starts with a 7-day free trial. All plans include access to Autopoetic's Humanist agent.
| Plan | Price | Typical usage | Work units / week |
|---|---|---|---|
| Basic | $5 / week | Light-to-moderate usage — 50,000 prompt + 50,000 completion tokens per week (soft limit) | 200 |
| Plus | $25 / week | 6× the usage of Basic — 300,000 + 300,000 tokens per week (soft limit) | 2,000 |
Signing up goes through Stripe Checkout — cancel or change plans any time from Settings → Manage Billing.
Limits and rate limiting¶
Weekly limits come in two tiers:
- Soft limit — you can keep going: new messages are accepted but rate-limited (processed at lower priority). The usage bar in the footer shows (rate limited) past this point.
- Hard limit — sending returns an error until the next weekly period starts, or until you upgrade: "You have exceeded your plan's hard limits for this period."
Work units¶
Work units measure workflow activity: every workflow step execution counts as one work unit against your weekly allowance. When the allowance is exhausted, workflow runs pause and resume automatically at the start of your next weekly period — no work is lost.
What counts as usage¶
Usage is metered against your plan per customer account, across all projects: chat messages (including perspective evaluations), Ask-mode questions, knowledge learning, project context updates, API requests, and workflow step executions. Hiding perspectives (chat features) reduces per-message usage at the same response quality.
Security¶
Your conversations are encrypted when you aren't using the service, but we do have to read your messages while you are chatting in order to respond to you. See projects and teams.